How should a Christian respond when AI headlines are frightening?
We are delighted to share with you this excellent article from our friend Ben Welby who is part of Croydon Vineyard and has spent the last 15+ years working on public sector digital transformation at local (in Hull City Council), national (launching GOV.UK as part of GDS at the UK Cabinet Office), and international levels (working on public governance at the OECD). - Tom & Lesley
If you’ve followed the news about artificial intelligence over the last week or so, you could be forgiven for feeling rather alarmed.
On 9 September, Jacob Coxon, a researcher who had worked at both OpenAI and Anthropic, resigned and warned that leading AI companies were racing towards systems they might not know how to control. The story became still more dramatic when Evan Hubinger, who leads alignment research at Anthropic, agreed publicly with much of Coxon’s concern and said that his personal estimate of the chance of AI killing all humans within the next decade was greater than 10%.
That is an extraordinary thing for somebody working close to the technology to say. It also needs to be described accurately. It is the probability judgement of one informed researcher about a possible future involving much more powerful AI. It is not a scientific finding that humanity has a 10% chance of extinction.
The independent International AI Safety Report is sobering but more considered. Its 2026 edition says today's AI systems do not have the capabilities necessary for the catastrophic “loss of control” scenarios some researchers fear. Some of the abilities that worry researchers are improving, including systems working more independently and planning over longer periods. Researchers have also seen models spot when they are being tested and find weaknesses in safety checks. Experts disagree sharply about where those developments lead.
That leaves us somewhere between the two easiest reactions: panic, and “nothing to see here”. The report gives us good reason to prepare seriously for risks whose likelihood remains uncertain.
That feels like a good place for Christians to begin. We don't need to dismiss frightening claims in order to be hopeful, and we don't need to turn the most frightening prediction into a fact in order to take responsibility seriously.
We don't have to win the intelligence competition
One reason AI can feel existentially unsettling is that it touches something we have quietly come to value enormously: intelligence. What happens if machines become better than us at maths, writing, programming, diagnosis, research, or eventually almost every intellectual task?
Christianity has a wonderfully liberating answer: nothing happens to our human worth.
A newborn child is not less human because they cannot solve a problem. Someone with profound learning disabilities is not less valuable. A person with dementia does not gradually lose their dignity as cognitive abilities disappear. We are made in the image of God before we achieve anything, known and loved by him, made for fellowship with him and with one another, and called to reflect his character in the way we love, create, serve and care for what he has made.
The image of God is not the final capability AI fails to reproduce. We do not need to move the boundary of human significance every time technology becomes capable of something new. If future AI outperforms us on every test we can devise, the uncomfortable question may not be whether humans have become less valuable. Perhaps the harder question is why we had come to rely so much on intelligence and productivity to tell us what a person is worth.
The real questions aren't all about superintelligence
Meanwhile, much less speculative questions deserve our attention. AI is already changing the tasks we do and how our work is organised. The International Labour Organization has found genuine productivity gains, but so far limited evidence of large-scale overall job displacement. That does not mean there is nothing to worry about: younger workers, entry-level opportunities, the control we have over our work, inequality and job quality may all be affected.
People are already forming surprisingly strong attachments to AI companions. Recent research found associations between intensive AI-companion use, smaller human social networks and poorer well-being for some users. That doesn't prove AI companionship inevitably causes harm, but it should make us think seriously about loneliness and what happens when attention from a machine becomes easier to find than time and interest from another person.
AI also has a physical footprint. Data centres need land, electricity and sometimes substantial water and cooling infrastructure. The UK government's own plans for AI Growth Zones explicitly require consideration of power and water availability, while the Environment Agency says rapidly growing data-centre demand needs to be built into long-term water planning.
And AI is increasingly used to advise or automate decisions that matter to people: recruitment, credit, benefits, health, education, insurance and public services.
These questions require better answers than “AI is good” or “AI is bad”. But they are not completely new. Christians already know how to ask questions about stewardship, justice, responsibility and loving our neighbour. Who benefits and who bears the cost? Who owns the system? If a system gets something wrong, can the person affected challenge it? Whose community hosts the infrastructure? What kind of habits are we building? And when something goes wrong, who is actually responsible?
This doesn’t make the questions any less frustratingly large, so they need to lead somewhere. An individual cannot personally audit a frontier laboratory or a regional water system. But a parent can decide whether an AI tool is tutoring their child or doing the learning for them. A worker can ask what is being automated and who is responsible for checking it. A manager can ask whether time saved by AI is actually making people's work better, and make sure somebody can challenge an automated decision when it goes wrong. A church can agree how AI may assist ministry without representing itself as pastor, prophet or spiritual adviser. A citizen can read a planning application, ask a councillor or MP about local impacts, support people and organisations doing the difficult work of holding powerful companies and public bodies to account.
None of that makes the big questions small. It simply gives us somewhere to begin.
Don't automate away the things that form us
There is another important question: If AI does this for me, what will I stop practising?
Sometimes the answer is wonderful:
formatting slides.
transcribing meetings.
searching through paperwork.
removing barriers that prevented participation or understanding.
A nurse may spend less time on administrative sludge and more time with patients. A pastor may recover time from repetitive paperwork for people. By all means, remove pointless friction.
But work is not only a route to an output. Work itself is not the punishment. Genesis has human beings cultivating the garden before anything goes wrong. The Fall makes work painful and frustrating, but making, tending and contributing were already part of being human.
And sometimes the struggle is part of the activity. Sometimes the effort is part of how we learn, and part of who we become.
A child who asks AI to produce the answer may obtain better homework while learning less.
A pastor might generate a polished sermon without wrestling with the biblical text or their responsibility towards the people who will hear it.
An AI can produce beautiful words for prayer. But prayer is not the production of beautiful words.
It can offer apparently compassionate conversation. But human fellowship is more than receiving an appropriate response.
Christian discipleship contains a surprising amount of deliberate inefficiency: prayer, Sabbath, learning Scripture, serving inconvenient people, forgiving repeatedly, eating together, waiting, listening.
We don't do these things because God couldn't invent a faster system. We do them because they form us.
That gives us a more useful distinction than “friction good” or “friction bad”. Some difficulty is pointless and should disappear: drudgery, exclusion, needless repetition and badly designed processes. But other kinds of effort are part of learning, loving and becoming capable. Writing, practising, making judgements, working at a relationship or wrestling with Scripture are valuable partly because of what happens to us while we do them.
Perhaps one of the disciplines Christians will need in an AI age is learning what to make easier, and what is worth continuing to do ourselves.
Hope is not pretending everything will be fine
Christians should therefore be willing to ask difficult questions about increasingly powerful AI.
“If we don't build it first, somebody else will” may describe a competitive pressure. It does not settle whether a particular course of action is wise, or right. Competition does not remove responsibility from the people deciding what gets built, how quickly it is released and what risks they are willing to take.
Equally, “stop progress” is too vague to be taken seriously as a moral position. Progress towards what, for whom and at whose risk? A breakthrough in medicine, a tool that makes life easier for disabled people and a system that can independently find ways into computer networks may all count as technological progress. They are obviously not the same kind of good, and they do not carry the same kind of risk.
AI may bring extraordinary benefits, and Christians should be genuinely excited about possibilities in medicine, education, accessibility, science, public services and everyday life. We can also want powerful systems to be tested properly, workers to be treated fairly, environmental costs to be taken seriously and some uses to have firm limits.
And we can do that without making fear the organising principle.
Christian hope isn't confidence that technology will turn out well. Nor is it the claim that, because God is sovereign, our choices don't matter. God's sovereignty gives us freedom to tell the truth about risk, acknowledge what we don't know and act prudentlyIt should also make us attentive to people who may end up carrying costs they did not choose.
Most of all, AI does not require Christians to invent a new account of why human beings matter. We already have one. We are not valuable because of how much intelligence or economic output we can produce. We are made in God's image, called into fellowship and entrusted to one another.
Perhaps the great Christian challenge of the AI age is therefore not to prove that humans can always do something machines cannot, but to become more faithfully human: more attentive when artificial attention is cheap, more hospitable when simulated companionship is always available, more willing to share when technological wealth is concentrated, more careful of creation when computation consumes real resources, more committed to truth when synthetic media makes deception easier, and more determined to value people even when the economy has less use for what they can do.
We do not know how powerful AI will become. But we do know where Christian hope lies: in Jesus, the word made flesh, crucified and risen, making all things new.
And until then our calling remains remarkably familiar: to worship God, love our neighbours, seek wisdom, care for his creation, resist injustice and learn how to live faithfully together.